We developed an analytical model which sought to explain the apparent cooperativity found in the binding of estrogen to its nuclear receptor by some groups. The current view is that the estrogen receptor binds bivalently, and possibly must first dimerize before it, together with DNA estrogen responsive elements and possibly other subunits, can bind steroid. The model was investigated numerically to find conditions where positive cooperativity might become apparent. We have found that while the model is sufficiently flexible to simulate cooperative behavior, most available data are not sufficiently precise as to rule out artifactual explanations for the observed cooperative behavior. Thus, curve-fitting of existing data is not yet feasible.